Gravity drainage processes are used for extracting highly viscous oil (“heavy oil”) from subterranean formations or bitumen from oil sand formations. For purposes of this patent specification, the general term “oil” will be used with reference to liquid petroleum substances recovered from subterranean formations, and is to be understood as including conventional crude oil, heavy oil, or bitumen, as the context may allow or require.
For heavy oil or bitumen to drain from a subterranean formation by gravity, its viscosity must first be reduced. The Steam-Assisted Gravity Drainage (SAGD) process uses steam to increase the temperature of the oil and thus reduce its viscosity. Other known gravity drainage processes use solvents or heat from in-situ combustion to reduce oil viscosity.
SAGD uses pairs of horizontal wells arranged such that one of the horizontal wells, called the producer, is located vertically below a second well, called an injector. The vertical distance between the injector and producer wells is typically 5 meters (5 m). The horizontal section of a SAGD well is typically 700 m to 1500 m long. For SAGD projects in the Athabasca oil sands in Alberta, Canada, the depth of the horizontal section is typically between 100 m and 500 m from the surface. Bitumen recovery from the oil sands is accomplished by injecting steam into the injector wellbore. Steam is injected from the injector wellbore into the hydrocarbon-bearing formation, typically through slots or other types of orifices in the injector wellbore liner. The steam permeates the formation within a region of the formation adjacent to the injector well; this steam-permeated region is referred to as a steam chamber. As steam is continuously injected into the formation, it migrates to the edges of the steam chamber and condenses at the interface between the steam chamber and the adjacent region of the bitumen-bearing formation. As the steam condenses, it transfers energy to the bitumen, increasing its temperature and thus decreasing its viscosity, ultimately to the stage where the bitumen becomes flowable, whereupon the mobile bitumen and condensed water flow down the edges of the steam chamber, accumulating as a “liquid inventory” in a lower region of the steam chamber and flowing into the producer wellbore. The fluid mixture of flowable bitumen and water that enters the producer well is then produced to the surface.
A significant challenge encountered by operators of SAGD well pairs is controlling the inflow distribution of oil and water over the horizontal length of the producer well, or the outflow distribution of steam, solvents, or combustion gases from the horizontal injector well. In many cases, inflow distributions or steam outflow distributions are biased towards one part of the well—for example, the region near the heel of the well (i.e., where the horizontal producer well transitions to a vertical well to the surface) or the region near the toe of the well. This results in less favourable well economics due to ineffective use of injection fluid (i.e., steam), poor bitumen recovery rates, and low recovery factors (i.e., when parts of the reservoir are not produced). The inflow/outflow biasing is influenced by the reservoir geology, which is largely outside the control of the well operator.
Another important factor influencing inflow and outflow distributions is the sand face pressure distribution along the length of the injector or producer well resulting from wellbore hydraulics. In this context, “sand face” refers to the point where flow emerges from the sand pack. In oil sands, the sand packs around the liner and flow emerges from the point where the sand is retained by the liner and flows into the gaps of the sand screen. The well operator has some control over this factor by means of the well completion design. For a typical injector well injecting steam into the formation through a slotted liner, wellbore steam pressures are highest near the heel and decrease towards the toe due to fluid friction pressure losses in the axial direction of the wellbore. Where wellbore pressures are higher at the heel, greater outflows of steam, solvent, or other injected gas are present. To equalize or create preferential outflow distributions, Dall'Acqua et al. have proposed (in International Application No. PCT/CA2008/000135) an injector completion with a tubing string run inside a liner, whereby the tubing string has ports located along its length that are sized and positioned to create a uniform or preferential sand face pressure distribution over the length of the injector well. The pressure distribution could be customized to achieve preferential outflow distributions into reservoirs with varying mobility (due to varying formation permeability, for example).
The experience of SAGD well operators in Alberta has shown that the performance of gravity drainage wells is affected by both injector and producer completion designs. In some cases, the producer completion has been shown to have a more significant effect on well performance. A method of controlling inflow distributions over the length of a long horizontal producer well is needed. Producer well design requires consideration of additional complexities that are not factors for injector well design. The fluid interface level relative to the producer needs to be managed carefully to both maximize production rates and to protect the producer well from breakthrough of injection gases. Breakthrough of steam into the producer will damage the well and/or related facilities, and breakthrough of other injection gases (e.g., light hydrocarbons such as propane and butane) reduces the efficiency of their function to mobilize bitumen.
The fluid interface (i.e., the interface between the liquid inventory and the overlying steam chamber) is characterized by a density contrast between the injection fluid (typically steam) and the produced oil and water. For purposes of this patent specification, the fluid interface level will be alternatively referred to as the “liquid level”. It is preferred to let the liquid level sit a short distance above the producer well to act as a seal preventing steam from entering the producer well. If steam is allowed to enter the producer, the steam is not being used for heating bitumen and the process becomes less efficient. Steam entering the producer well can also carry sand particles at high speeds and cause erosion of the steel liners and tubing strings in the wellbore.
To evaluate the economics of an oil recovery project, an estimate of the recovery rate is required. For conventional oil wells, an inflow performance relationship (IPR) is used to predict the oil recovery rate for the reservoir pressure and bottom hole pressure conditions expected. In this sense, conventional oil production is driven by pressure not gravity. Therefore, IPRs as used for conventional oil wells cannot be applied to gravity drainage projects, so a gravity drainage inflow performance relationship (GIPR) is needed to estimate the economics of the process.
“Thermal Recovery of Oil and Bitumen” (R. Butler, 1997, 3rd edition, printed by GravDrain Inc., ISBN 0-9682563-0-9) presents formulas for predicting SAGD recovery rates for a given liquid head, or difference in height between the top of the steam chamber and the producer well. The calculation is based on a two-dimensional cross-section of the well and reservoir. Two other factors will affect SAGD production rates that are not covered in these calculations. Firstly, Butler's calculation assumes that the liquid level contacts the top of the producer well. In actuality, it is typical for liquid levels to sit above the producer wellbore forming a liquid “trap” that the producer wellbore is submersed in. As bitumen and water flow through the liquid trap to the producer well, pressure loss will occur. Many SAGD operators have observed significant pressure losses in this region, with resultant reduction in actual production rates relative to predicted rates. While exact causes for these pressure losses are not fully known, they are sometime attributed to two-phase flow (relative permeability) effects, plugging of slotted liners, fines migration, or other causes.
Another important consideration for predicting SAGD production rates is that wellbore pressures and temperatures vary along the length of a long horizontal well. This will cause liquid levels, and thus the depth of the liquid trap, to also vary along the length of the well, which in turn will affect the total production rate from the well. Near-wellbore reservoir heterogeneities (i.e., permeability variations close to the wellbore) will also contribute to inflow variations along the length of the well.